Introduction

Sight is a critical sense for everyone and understanding a child’s visual abilities and limitations is vital to optimally support their development [1]. There is a body of evidence that children with learning disabilities are significantly more likely to experience eye and vision problems than members of the general childhood population [2–7]. Children with the most complex needs are likely to be taught in special schools. In England 79% of children with severe learning disabilities and 81% of children with profound and multiple learning disabilities attend such schools [8]. The total population attending special schools in England at the time of writing is estimated at 117,888 [9].

Vision screening in the general childhood population improves the identification and correction of reduced vision and is implemented in many countries worldwide [10]. Under the recommendations of the United Kingdom National Screening Committee [11], all children should have access to an orthoptic led screening of vision at school entry (age 4–5 years) primarily to identify amblyopia, but also refractive error and strabismus. However, it has been suggested that the provision of this type of screening, in special schools is patchy [5] and it is questionable if such screening is suitable for children with complex needs, given that many cannot be assessed using established visual acuity tools [4]. Studies [2,4,5] have also highlighted the low uptake of hospital eye care and primary care optometric eye examinations in community optical practices by the special schools population. The UK eye care professional bodies with SeeAbility have developed a recommended Framework for special schools eye care [12] which advocates the provision of routine eye care, including refraction, eye health checks and spectacle dispensing, in the special school environment. In 2017, Public Health England made resources available to support the commissioning and delivery of child vision screening and these recommend a routine full eye examination should be offered in school for all children attending a special school [13].

This study reports on service evaluation data from an opt-in school based eye care service provided by optometrists, orthoptists and dispensing opticians working with SeeAbility, a UK charity supporting people with learning disability and sight loss, across 11 special schools in England between 2013 and 2017. This evaluation sets out to report on uptake of the service, clinical findings, and self-reported eye care history of children attending at first appointment.

Methods

Eye examinations were offered to children attending 11 special schools in London (8 schools), Buckinghamshire, Durham and Manchester. Informed parental consent (see S1 Fig.) was obtained prior to the eye care team visiting the school and parents were given a detailed questionnaire exploring previous ocular and general health history prior to the first visit [14],(see S2 Fig).

The questionnaire included 5 questions used to identify features of cerebral visual impairment (CVI) [15]. Cerebral visual impairment is an umbrella term for visual perceptual difficulties with a cerebral, rather than ocular origin. The cerebral deficit is often unknown, but manifestations of CVI are more common amongst children with a history of brain injury or maldevelopment. CVI can range from obvious poor sight to difficulties in specific tasks (such as negotiating steps) with good visual acuity. Children with a score of 3 or more were designated as suspect CVI and referred to a qualified teacher for the visually impaired (QTVI) for a more detailed inventory of questions where this service was available. Some children in the study had a diagnosis of CVI reported by parents or school; these were all children with recognisable poor sight.

In each case primary general health diagnosis was recorded as reported by parents or, when available, from school medical records.

During the 4 years since the service was launched, eye examinations were carried out by 6 optometrists, all experienced in working with children with disabilities. Examinations took place in suitable rooms of minimum 3 metres in length which could be blacked out and children were accompanied to the test by a member of teaching staff who knew them well and/or a parent. Refraction was conducted by retinoscopy, with the use of 1% cyclopentolate drops to produce cycloplegia when clinically necessary and practically possible in the judgement of the optometrist. Visual acuity was measured using available age and ability-appropriate tests. Tests available included Keeler Preferential Looking Cards [16] Cardiff Acuity Test[17], Bradford Visual Function Box[18] (in year 4 of operating our service), LogMAR linear and single crowded versions of Kay Pictures[19] and Keeler LogMAR Crowded letter test (formerly Glasgow Acuity Cards) [20].

For children starting reception year (at age 4–5) or nursery, a joint optometric/orthoptic assessment was carried out when possible, with visual acuity and binocular vision status being measured by an orthoptist. For all children, we attempted to measure vision and/or visual acuity monocularly and binocularly, with the pupil’s existing spectacles if worn. For all children, we attempted an assessment of binocular vision using tests including observation of corneal reflections, cover tests at distance and near, assessment of ocular motility, near point of convergence, 20/10 base out prism test, observation of smooth pursuit and saccadic eye movement ability, quality of fixation, stereopsis, and observation of visual reflexes (Optokinetic nystagmus, vestibulo-ocular reflex, blink). Pupil reactions, media quality, and an assessment of internal and external ocular health were attempted for all pupils.

Spectacles were prescribed when required. Parents or carers were given the choice of a spectacle dispensing service offered in school or being issued with a spectacle prescription voucher to use at a community optical practice.

Pupils were referred to their GP or for an ophthalmological opinion in secondary care according to clinical judgement.

A detailed lay language report form designed in collaboration with Ulster University School of Optometry and Vision Science, was completed for each child (S3 Fig)[21](. This was designed to clearly explain the findings of the assessment including the child’s visual abilities, recommended actions (including need for spectacles wear when necessary) and ongoing care plan. The report was designed to be easily incorporated into each child’s Education Health and Care Plan (EHCP); it was given to parents and school and was made available to Qualified Teachers of the Visually Impaired (QTVIs) when visual defects were identified. Reports were also shared when relevant and, depending on local arrangements with school paediatricians, speech and language teams and occupational therapists.

The data for this service evaluation was derived from annual audit of the SeeAbility service, which does not require ethical approval. Nevertheless, the proposed use of audit data was scrutinised and approved by the Cardiff School of Optometry and Vision Sciences Human Research Ethical Committee, prior to commencement of the service.

Our analysis of data was carried out using SPSS (IBM Corp., version 20.0.0.2) and was limited to the first examination of each pupil; this may have been performed over more than one contact during a period of a few weeks to gain as complete a clinical assessment as possible. Where children were absent, unwell or unable to complete their assessment, it was possible to rearrange appointments to allow maximally efficient use of clinician’s time.

The distribution of refractive error is influenced by ethnicity [22] and the schools covered by the SeeAbility service have a varied ethnic mix. The ethnicity of individual pupils participating in the service was not recorded at the time of testing. However, the ethnic diversity of the London schools is available, and figures for September 2016 across the eight participating London schools were: White European 34%, South Asian 12%, Black African Caribbean 32% and ‘other’ (East Asian, mixed or minority groups such as Afghan, Turkish etc.) 22%.

Anecdotal clinical observation indicated pupils with a diagnosis of autism were less likely to provide a reliable visual acuity measurement using the available tests than their peers with a learning disability of a different aetiology. We explored this in our data analysis by comparing ability to obtain a reliable visual acuity measurement in pupils with and without an autism diagnosis.

Results

Over the four-year period 949 new children were seen; 509 were of primary school age and 440 secondary school age. Mean age was 10.7 years (range was 3.0 to 19.8 years).

Clinical findings

Reported history of previous eye care.

Table 1 shows the parent reported eye care history for pupils seen by the service. This information was available for 847 pupils (89.3%).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Parent reported eye care history.

https://doi.org/10.1371/journal.pone.0212733.t001

Of pupils for whom history was available 373 (44%) reported no previous eye care. Overall, 91 (10.7%) reported having attended a community optical practice. A further 6 pupils (0.7%) had been seen in school by an optometrist under special arrangements in which children were referred by a paediatrician.

Presenting level of vision.

It was possible to assess every child for whom consent was received but often with only limited engagement in some elements of formal testing. In some cases, repeat attempts at assessment were necessary, or it was necessary to see children in their own classroom, a ball pool, corridor or sensory room.

A reliable measure of presenting visual acuity using a formal test was recorded for 574 (60.5%) pupils. This was defined as the acuity in the better eye or the binocular acuity when no monocular acuity was available. Low vision (WHO criterion [23] of LogMAR 0.5 or poorer in the better eye, or binocular) was recorded for 143 pupils (24.9% of the pupils for whom it was possible to record a presenting level of vision). This included 26 pupils with severe visual impairment and 12 pupils with CVI that precluded quantitative acuity measurement.

The recommended pass criteria for UK school vision screening for ages 4–5 years [13] is a monocular score of 0.2 LogMAR using the Keeler LogMAR crowded letter test. In the SeeAbility service, 156 pupils fell into this age group (4.00 to 5.97 years) but only 10 pupils were able to use a LogMAR crowded acuity type test (Kays Pictures LogMAR or Keeler LogMAR) and only one (0.6%) was able to pass the test using the recommended Keeler test.

Of the 574 pupils able to take part in acuity assessment and accepting binocular acuity in cases in which no monocular acuity was available, 300 (52.3%) pupils failed to achieve LogMAR 0.2 using any of the tests available. Using the more rigorous criterion of successful monocular acuity recording of LogMAR 0.2 or better in each eye, a total of 76.8% of special school pupils across all ages would fail vision screening.

Of the 375 pupils with no acuity score recorded, the reason for this was recorded for 348 pupils. Of these, 38 pupils had severe visual impairment or CVI and for 310 pupils co-operation / comprehension was too limited to achieve a reliable threshold result.

A diagnosis of disability was available for 855 pupils. For the purposes of analysis of ability to co-operate for visual acuity, the pupils were divided into those with a diagnosis of autism (444, 51.9%), and those with another diagnosis (such as Down’s syndrome, cerebral palsy, global developmental delay, specific syndromes etc., 411, 48.1%). Some children had a dual diagnosis of Down’s syndrome and autism and were categorized under autism. Pupils whose acuity was too poor to attempt acuity were excluded. Among the pupils who had autism, 61.7% were cooperative for acuity, while among pupils without autism, 68.9% were cooperative for acuity testing. The difference was significant (χ2 = 4.41, p = 0.04); children with autism were less likely to be successfully tested for acuity.

Incidence and type of refractive error requiring spectacle correction.

An assessment of the level of refractive error was possible for 932 (98.2%) right eyes and 928 (97.8%) left eyes using the objective method of retinoscopy. Cyclopentolate 1% eye drops were used for 47 (5.0%) examinations. (Cyclopentolate is routinely used in paediatric ophthalmic assessment to measure refractive error by paralysing accommodation and to gain a good retinal view through a dilated pupil).

Table 2 shows the spectacle prescribing pattern of the service.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Spectacle prescribing pattern.

https://doi.org/10.1371/journal.pone.0212733.t002

Overall 646 (68.1%) pupils had either no refractive error or very low refractive error which did not indicate the need for spectacles; 299 (31.5%) children needed spectacles and for 122 (12.9%) these were prescribed for the first time, A very high proportion of the 299 parents /carers of children needing spectacles (n = 288, 96.3%) chose to have them dispensed and fitted in school.

Myopia of ≤-0.50 in either eye was measured in 212 pupils (22.3%); 38 pupils (4.0%) were high myopes (<-6.00D), including 8 children with myopia in excess of -13.00DS. The high myopes comprised 19 of primary school age and 19 of secondary age. Of this group of high myopes, 6 children were given spectacles for the first time by the service.

Hypermetropia of ≥+2.00 in either eye was measured in 144 (15.2%) pupils. Astigmatism was defined as ≥1.00DC in either eye and was noted for 271 (28.6%) pupils (astigmatism can be present in isolation or co-exist with myopia or hyperopia).

Accommodative accuracy is the ability to accurately focus onto a close target when corrected for any hyperopia, myopia or astigmatism and it is often impaired in children with learning disabilities [24,25,26]. This was assessed in 700 pupils: of these 572 (81.7%) had accurate accommodation and 74 (10.6%) had inaccurate accommodation. The children with under accommodation were prescribed spectacles incorporating a prescription for near vision, for 100% of these pupils, this was for the first time.

Incidence of strabismus and other problems with eye movement.

A problem with eye movements/alignment (an orthoptic anomaly) was recorded in 239 (25.2%) of pupils; 214 (22.6%) had strabismus and 23 (2.4%) had nystagmus (involuntary eye movements). Other orthoptic anomalies recorded included poor eye movement control and lack of smooth pursuit eye movements.

Incidence of other pathology.

A pathological condition was recorded in 72 (7.6%) pupils; some pupils had more than one condition. This included 4 cases of blepharitis (eyelash infection), 5 cases of conjunctivitis, 4 cases of keratoconus, 12 cases of cataract, 4 optic disc anomalies and 7 retinal anomalies.

Visual field defects and cerebral visual impairment (CVI).

Hemianopia is a total loss of one half of the vision in each eye, originating at a neural level; 13 cases were recorded, 5 of which were previously undiagnosed. A further 10 pupils had another gross visual field defect.

Cerebral visual impairment, a problem with visual processing at a cerebral level, was diagnosed or suspected in 141 (14.9%) pupils. All were referred to the QTVI for further evaluation.

Incidence of visual deficiency of any aetiology.

A visual deficiency was recorded when a child had a defect of any aetiology (correctable refractive error, orthoptic anomaly, ocular pathology or suspected / confirmed cerebral visual impairment) irrespective of visual acuity scores. Results were available for 923 (97.3%) children; a decision would not be made for the remaining pupils after first assessment. Of these 923 pupils for whom it was possible to confirm the presence or absence of a visual deficiency 438 (47.5%) pupils had a visual deficiency and 119 of these (12.9%) were previously undiagnosed.

Of the 373 pupils with no history of previous eye care reported, 61 (16.4%) had a visual deficiency that was not previously known, 47 (12.6%) needed a spectacle prescription, 37 (9.9%) had low vision according to WHO criteria [23] including 5 pupils who had vision too low to be recorded (1.3%).

Onward referral.

Twelve (1.3%) pupils were referred to their GP for further management and 17 (1.8%) for the opinion of an ophthalmologist. In addition 32 (3.4%) pupils were referred by the optometrist for an orthoptic opinion—this was provided in school by a SeeAbility orthoptist (this number discounts the new starters who received an orthoptic assessment as part of the service).

Discussion

This evaluation shows that many children in our special school population have undiagnosed visual needs. Our findings align with those of previously published studies which reported the prevalence of ocular and visual anomalies among children with learning disabilities, including those in special schools elsewhere in the UK [2,4,5] and in other countries [3,6,7].

Previously published work has looked at children attending special schools in Wales [5], Scotland [2] and the north of England [4]. In the present study, 87.0% of pupils attended schools in the south east of England, mainly in the greater London area; thus it helps to inform the UK wide picture.

Limitations

The study did not explore the eye care history or status of around 20% of pupils who did not opt into the service in the schools where the service was offered. These findings cannot therefore be taken to represent the entire special school population.

Conclusion

The very low uptake of free NHS eye examinations in our population suggests that the NHS General Ophthalmic Service is currently failing children with complex needs in England. Furthermore, only one child of 156 of school entry age would have passed routine vision screening as currently recommended by the National Screening Committee [11]. Routine vision screening is demonstrably inappropriate for this population. This and other UKstudies [2,4,5], provide evidence of a high incidence of visual deficiencies, including a high incidence of correctable refractive error, in the UK special school population. The study provides evidence that existing eye care services in England are failing to identify and manage a significant proportion of children with visual deficiencies This is evidence that children attending special schools need a targeted in-school eye care service that provides full, regular, routine, eye examinations and spectacle provision as described by the Framework for provision of eye care in special schools in England [12] and recommended by Public Health England’s screening guidelines [13].

Supporting information

Acknowledgments

The authors wish to thank Raimonda Bullaj, Tania Stewart-Davis, Simon Berry and Daniel Crown, optometrists, Laura Smith, Miriam Conway, Rhiannon Bader and Helen Downey, orthoptists and Edward Saunders, Dispensing Optician for their part in data collection and the SeeAbility administrative team. Thanks also to the school nurses, administrators and teaching staff of the special schools involved (S4 Fig) for their contribution to the success of the eye examinations.